US20030095692A1 - Method and system for lung disease detection - Google Patents

Method and system for lung disease detection Download PDF

Info

Publication number
US20030095692A1
US20030095692A1 US09683111 US68311101A US2003095692A1 US 20030095692 A1 US20030095692 A1 US 20030095692A1 US 09683111 US09683111 US 09683111 US 68311101 A US68311101 A US 68311101A US 2003095692 A1 US2003095692 A1 US 2003095692A1
Authority
US
Grant status
Application
Patent type
Prior art keywords
anatomical
models
method
system
information
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
US09683111
Other versions
US7058210B2 (en )
Inventor
Joseph Mundy
Colin McCulloch
Ricardo Avila
Shannon Hastings
Robert Kaucic
William Lorensen
Matthew Turek
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
General Electric Co
Original Assignee
General Electric Co
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date

Links

Images

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING; COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00Image analysis
    • G06T7/0002Inspection of images, e.g. flaw detection
    • G06T7/0012Biomedical image inspection
    • AHUMAN NECESSITIES
    • A61MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61BDIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B6/00Apparatus for radiation diagnosis, e.g. combined with radiation therapy equipment
    • GPHYSICS
    • G06COMPUTING; CALCULATING; COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00Indexing scheme for image analysis or image enhancement
    • G06T2207/30Subject of image; Context of image processing
    • G06T2207/30004Biomedical image processing
    • G06T2207/30061Lung

Abstract

A method for processing medical images for use in the detection and diagnosis of disease comprises classifying regions of interest within the medical images based on a hierarchy of anatomical models and signal models of signal information of an image acquisition device used to acquire the medical images. The anatomical models are derived to be representative of anatomical information indicative of a given disease. A computer-aided system for use in the diagnosis and detection of disease comprises an image acquisition device for acquiring a plurality of image data sets and a processor adapted to process the image data sets. The processor is adapted to classify selected tissue types within the image data sets based on a hierarchy of signal and anatomical models and the processor is further adapted to differentiate anatomical context of the classified tissue types for use in the diagnosis and detection of disease.

Description

    BACKGROUND OF INVENTION
  • This invention relates to a method and system for processing medical image data to aid in the detection and diagnosis of disease, and more particularly, to a method and system for detecting lung disease in medical images obtained from a x-ray computed tomography (CT) system. [0001]
  • A x-ray chest radiograph system is the more commonly used diagnostic tool useful for the purpose of detecting lung disease in humans. Lung disease such as bronchitis, emphesema and lung cancer are also detectable in chest radiographs and CT. However, CT systems generally provide over 80 separate images for a single CT scan thereby providing a considerable amount of information to a radiologist for use in interpreting the images and detecting suspect regions that may indicate disease. [0002]
  • Suspect regions are defined as those regions a trained radiologist would recommend following through subsequent diagnostic imaging, biopsy, functional lung testing, or other methods. The considerable volume of data presented by a single CT scan presents a time-consuming process for radiologists. Conventional lung cancer screening generally involves a manual interpretation of the 80 or more images by the radiologist. Fatigue is therefore a significant factor affecting sensitivity and specificity of the human reading. In other diseases, such as emphysema, it is difficult for a radiologist to classify the extent of disease progression by only looking at the CT images. Quantitative analysis of the anatomy is required. [0003]
  • Attempts to automate lung cancer and emphysema detection in CT scans have been based on a variety of nodule detection and classification techniques, and lung parenchyma metrics. The emerging field is referred to as Computer Aided Diagnosis, or alternatively, Computer Aided Detection (CAD). There is a significant amount of literature on methods for automating lung cancer detection in CT scans. Generally nodule detection has proceeded in three steps: lung segmentation, vessel extraction, and final nodule candidate detection and classification. [0004]
  • Vessel extraction has been attempted using gray-level thresholding, fuzzy clustering, and three-dimensional seeded region growing). Nodule detection has been done using template matching, genetic algorithms, gray-level thresholding, the N-Quoit filter, region growing, and edge-gradient techniques. [0005]
  • Once candidate nodules are produced by any of the above methods, classification has been implemented via rule-based methods, neural network classification, fuzzy logic, and statistical techniques including factor analysis and linear discriminating analysis. [0006]
  • The above techniques presented to date, however, have largely focused on identifying suspicious lesions in CT scans and have not directly addressed obtaining correct differentiation of structures in the lung and correct measurements of their size. Additionally, the above techniques are generally limited in the interpretative nature of the results. Typically, identification and classification of a lesion using the above techniques may produce a positive affirmation of a nodule, but further radiologist qualitative review and interpretation of results is generally required. For example, radiologists rely heavily on their familiarity with or expert knowledge of pathological and anatomical characteristics of various abnormal and normal structures in interpreting medical images. Further, the characteristics of the scanning device, such as type, pixel intensity and signal impulse response, also influence the presentation of the image data. A radiologist's interpretation of medical images also generally relies on his or her familiarity with a given scanner. There has been no apparent evaluation by the above techniques to address the type of or characteristics of the scanning device in the analysis of the images produced. [0007]
  • What is needed is a robust method and system for processing image data to produce quantitative data to be used in detecting disease. What is further needed is a method and system that provides interpretative results based on expert knowledge of a disease as well as the scanner capabilities and characteristics. Additionally, there is a requirement for the ability to track a disease's progression/regression resulting from drug therapy. [0008]
  • SUMMARY OF INVENTION
  • In a first aspect, a method for processing medical images for use in the detection and diagnosis of disease is provided. The method comprises classifying regions of interest within the medical images based on a hierarchy of anatomical models and signal models of signal information of an image acquisition device used to acquire the medical images. The anatomical models are derived to be representative of anatomical information indicative of a given disease. [0009]
  • In a second aspect, a computer-aided system for use in the diagnosis and detection of disease is provided. The system comprises an image acquisition device for acquiring a plurality of image data sets and a processor adapted to process the image data sets. The processor is adapted to classify selected tissue types within the image data sets based on a hierarchy of signal and anatomical models and the processor is further adapted to differentiate anatomical context of the classified tissue types for use in the diagnosis and detection of disease.[0010]
  • BRIEF DESCRIPTION OF DRAWINGS
  • The features and advantages of the present invention will become apparent from the following detailed description of the invention when read with the accompanying drawings in which: [0011]
  • FIG. 1 is a block diagram illustration of a medical imaging system for which embodiments of the present invention are applicable; [0012]
  • FIG. 2 is a flow diagram of a method for processing image data for use in detecting disease in accordance with embodiments of the present invention; [0013]
  • FIG. 3 is a flow diagram of a segmentation method useful in the medical imaging system of FIG. 1; and, [0014]
  • FIG. 4 is a block diagram illustration of a modeling method for use in detecting disease in accordance with embodiments of the present invention. [0015]
  • DETAILED DESCRIPTION
  • Referring to FIG. 1, a general block diagram of a system [0016] 100 for disease detection is shown. System 100 includes an imaging device 110, which can be selected from a number of medical imaging devices known in the art for generating a plurality of images. Most commonly, computed tomography (CT) and magnetic resonance imaging (MRI) systems are used to generate a plurality of medical images.
  • During a CT imaging session, a patient lies horizontal and is exposed to a plurality of x-rays measured with a series of X-ray detectors. A beam of x-rays passes through a particular thin cross-section or “slice” of the patient. The detectors measure the amount of transmitted radiation. This information is used to compute the x-ray attention coefficient for sample points in the body. A gray scale image is then constructed based upon the calculated x-ray attenuation coefficients. The shades of gray in the image contrast the amount of x-ray absorption of every point within the slice. The slices obtained during a CT session can be reconstructed to provide an anatomically correct representation of the area of interest within the body that has been exposed to the x-rays. [0017]
  • During a MR imaging session, the patient is placed inside a strong magnetic field generated by a large magnet. Magnetized protons within the patient, such as hydrogen atoms, align with the magnetic field produced by the magnet. A particular slice of the patient is exposed to radio waves that create an oscillating magnetic field perpendicular to the main magnetic field. The slices can be taken in any plane chosen by the physician or technician (hereinafter the “operator”) performing the imaging session. The protons in the patient's body first absorb the radio waves and then emit the waves by moving out of alignment with the field. As the protons return to their original state (before excitation), diagnostic images based upon the waves emitted by the patient's body are created. Like CT image slices, MR image slices can be reconstructed to provide an overall picture of the body area of interest. Parts of the body that produce a high signal are displayed as white in an MR image, while those with the lowest signals are displayed as black. Other body parts that have varying signal intensities between high and low are displayed as some shade of gray. [0018]
  • Once initial MR or CT images have been obtained, the images are generally segmented. The segmentation process classifies the pixels or voxels of an image into a certain number of classes that are homogeneous with respect to some characteristic (i.e. intensity, texture, etc.). For example, in a segmented image of the brain, the material of the brain can be categorized into three classes: gray matter, white matter, and cerebrospinal fluid. Individual colors can be used to mark regions of each class after the segmentation has been completed. Once the segmented image is developed, surgeons can use the segmented images to plan surgical techniques. [0019]
  • Generally, creating a segmented CT or MR image involves several steps. A data set is created by capturing CT or MR slices of data. Through the segmentation process, a gray scale value is then assigned to each point in the data set and different types of tissues will have different gray scale values. Each type of material in the data is assigned a specific value and, therefore, each occurrence of that material has the same gray scale value. For example, all occurrences of bone in a particular image may appear in a particular shade of light gray. This standard of coloring allows the individual viewing the image to easily understand the objects being represented in the images. [0020]
  • FIG. 1 illustrates a medical imaging system [0021] 100 to which embodiments of the invention are applicable. The system includes an imaging device 110, a processor 120 and an interface unit 130. Imaging device 110 is adapted to generate a plurality of image data sets 240 and is, for example, a computed tomography (CT) or magnetic resonance (MR) scanner. In the context of CT or MR, acquisition of image data is generally referred to as “scans”. Processor 120 is configured to perform computations in accordance with embodiments of the present invention which will be described in greater detail with reference to FIGS. 2-4. Processor 120 is also configured to perform computation and control functions for well-known image processing techniques such as reconstruction, image data memory storage, segmentation and the like. Processor 120 may comprise a central processing unit (CPU) such as a single integrated circuit, such as a microprocessor, or may comprise any suitable number of integrated circuit devices and/or circuit boards working in cooperation to accomplish the functions of a central processing unit. Processor 120 desirably includes memory. Memory within processor 120 may comprise any type of memory known to those skilled in the art. This includes Dynamic Random Access Memory (DRAM), Static RAM (SRAM), flash memory, cache memory, etc. While not explicitly shown in FIG. 1, the memory may be a single type of memory component or may be composed of many different types of memory components. Processor 120 is also capable of executing the programs contained in memory and acting in response to those programs or other activities that may occur in the course of image acquisition and image viewing. As used herein, “adapted to”, “configured” and the like refer to mechanical or structural connections between elements to allow the elements to cooperate to provide a described effect; these terms also refer to operation capabilities of electrical elements such as analog or digital computers or application specific devices (such as an application specific integrated circuit (ASIC)) that are programmed to perform a sequel to provide an output in response to given input signals.
  • Interface unit [0022] 130 is coupled to processor 120 and is adapted to allow human users to communicate with system 100. Processor 120 is further adapted to perform computations that are transmitted to interface unit 130 in a coherent manner such that a human user is capable of interpreting the transmitted information. Transmitted information may include images in 2D or 3D, color and gray scale images, and text messages regarding diagnosis and detection information. Interface unit 130 may be a personal computer, an image work station, a hand held image display unit or any convention image display platform generally grouped as part of a CT or MRI system.
  • All data gathered from multiple scans of the patient is to be considered one data set. Each data set can be broken up into smaller units, either pixels or voxels. When the data set is two-dimensional, the image is made up of units called pixels. A pixel is a point in two-dimensional space that can be referenced using two dimensional coordinates, usually x and y. Each pixel in an image is surrounded by eight other pixels, the nine pixels forming a three-by-three square. These eight other pixels, which surround the center pixel, are considered the eight-connected neighbors of the center pixel. When the data set is three-dimensional, the image is displayed in units called voxels. A voxel is a point in three-dimensional space that can be referenced using three-dimensional coordinates, usually x, y and z. Each voxel is surrounded by twenty-six other voxels. These twenty-six voxels can be considered the twenty-six connected neighbors of the original voxel. [0023]
  • In an embodiment of the present invention, a computer-aided system for use in the diagnosis and detection of disease comprises an image acquisition device for acquiring a plurality of image data sets and a processor adapted to classify selected tissue types within the image data sets based on a hierarchy of signal and anatomical models. The processor is further adapted to differentiate anatomical context of the classified tissue types for use in the diagnosis and detection of a selected disease. The system further comprises an interface unit for presenting the classified tissue types within the image data sets and anatomical context of the classified tissue types for aiding an interpretation of the processed image data sets. The anatomical models are parametric, mathematical representations of anatomical tissues. The anatomical context comprises at least one of lung nodules indicative of lung cancer, healthy lung tissue, diseased lung tissue indicative of chronic obstructive pulmonary disease (COPD) and other pathological descriptions of tissue that can be characterized by radiologists and further modeled mathematically. Further discussion of anatomical context and mathematical modeling will be provided with reference to FIG. 4. [0024]
  • In an exemplary embodiment, the imaging device is a x-ray CT scanner. A CT system is particularly well adapted to acquire a plurality of images, or alternatively slices, of a region of interest. Also, in this exemplary embodiment, the imaging object is a lung. It is to be appreciated that other imaging devices that provide a plurality of images, such as magnetic resonance (MR), would also benefit from embodiments of the present invention. Also, it is to be appreciated that other regions of interest other than the lung may be the imaging object, e.g. the heart, colon, limbs, breast or brain. The processing functions performed by processor [0025] 120 would be adapted to classify tissue types of interest in these other imaging objects.
  • An embodiment for a method for detecting disease from the plurality of medical images comprises the steps of acquiring the image data, processing of the acquired image data to define the lung region; computing low level features in the image using the known characteristics of the imaging device and the imaging process; grouping regions in the image, based on their features and an information object hierarchy describing their features, into anatomical structures; and, deciding if any of the grouped regions represents an area which is suspicious for a lung disease. The method further comprises presenting the areas identified as suspicious for lung disease. The presenting step comprises presenting the anatomical context (e.g. lung nodule, diseased tissue, healthy tissue) and a decision process by which the suspicious areas were identified. The grouping of regions is performed using comparisons of signal and anatomical models using Bayes Factors. In a further embodiment, a method for characterizing tissue in medical images for use in disease diagnosis and detection comprises computing an information object hierarchy of increasing complexity to characterize anatomical tissue. The object hierarchy contains models, or alternatively mathematical representations, based on characteristics of an image acquisition device used in acquiring the images and based on anatomical characteristics of a selected region of interest and a specified disease. The grouping, the object hierarchy and Bayes Factor comparisons will be described in greater detail in paragraphs that follow and with reference to FIG. 4. [0026]
  • Referring to FIG. 2, there is shown a more detailed flow diagram of an embodiment of a method for processing image data to be used in detecting disease. Image data is acquired at [0027] 210. These images are passed to processor 120 (FIG. 1) for processing steps 220 280 of FIG. 2. At step 220, the area of the images that represents the lung is determined by selection of various known segmentation techniques or, alternatively by an exemplary embodiment of pleural space segmentation which will be discussed in greater detail below with reference to FIG. 3. Resulting from step 220, input pixels from a CT scan are first classified to be either in the lung cavity or outside the lung. The input pixels are acquired from either a two-dimensional CT scan data set or, alternatively, from a three-dimensional CT scan data set. At 230, processor 120 then computes low-level signal models from the gray scale values of the image within the lung region. These models, for example, may include (but are not limited to) compact, bright objects; compact, dark objects; and long, bright objects. The low-level signal models are mathematical descriptions of structures being imaged after the measurement process of the scanner modifies them. Signal model processing continues at 250 to gain more information regarding a region of pixels in the image. In an embodiment of signal model processing for step 250, different signal models are competed against each other in order to best explain a region of pixels in the images. The competition is desirably carried out by performing comparisons between the signal models using the known statistical-based process of Bayes Factors. It is to be appreciated that other decision or statistical based methods may also be used. An exemplary embodiment using Bayes Factors will be described in greater detail below and with reference to FIG. 4.
  • After decisions have been made regarding the best low-level signal model, a further grouping process occurs at steps [0028] 260 and 270. This involves grouping the low-level models into anatomical structures such as particular areas of the lung. Again, the decision process involves competing anatomical models desirably using Bayes Factors in order to make an optimal decision as to model applicability.
  • Finally, at step [0029] 280, results are presented. Results are based on the information provided by the low-level signal models and the anatomical models in order to provide qualitative and quantitative information regarding suspicion for lung disease. Decisions at this level are made in the same way that a radiologist might make decisions regarding a lung nodule because the system has both low-level signal knowledge and anatomical context.
  • Referring to FIG. 3, an embodiment for identifying the lung region at step [0030] 220 is provided. In this embodiment, a lung segmentation process is provided that automatically identifies the boundaries of the pleural space in a Computed Tomography (CT) data set. The boundary is either a set of two-dimensional (2D) contours in a slice plane or a three-dimensional (3D) triangular surface that covers the entire volume of the pleural space. The extracted boundary can be subsequently used to restrict Computer Aided Detection (CAD) techniques to the pleural space. This will reduce the number of false positives that occur when a lung nodule detection technique is used outside the pleural space.
  • Referring further to FIG. 3, the 3D surface identification proceeds as follows: [0031]
  • [0032] 310 Acquire a CT data set that covers the lung. The extent of the CT exam should cover the entire region of the pleural space. The centerline landmark of the exam should run approximately down the center of the thorax.
  • [0033] 311 Read the CT data set into memory. For efficiency, the data set should reside in contiguous memory, although others means of memory organization are possible.
  • [0034] 312 Select a threshold. Select an intensity value that corresponds approximately to air in the CT study. This intensity is called the threshold. The threshold can be chosen using a variety of means, but only needs to be done once per CT lung protocol. The same threshold can be used for all exams using the same protocol (e.g. scanning procedure).
  • [0035] 313 Segment the study into foreground and background regions. Replace all samples that have values below the threshold with a positive constant foreground value. Replace all other samples with a 0, the background value. The actual foreground value is arbitrary. Samples marked with the foreground value will correspond to air while samples with a background value will correspond to other tissue.
  • [0036] 314 Remove islands in the xy, xz and yz planes. Islands are groups of samples that contain 0 but are surrounded by non-zero samples. Islands are removed by setting their values to the foreground value. Only islands that are below a specified island size are removed. The island size is chosen to be larger than the area of the cross-section of a vessel or bronchial passage and smaller than the area of background outside the CT circle of reconstruction.
  • [0037] 315 Select a seed in the pleural space. The seed is located in the middle slice, one quarter of the distance from the left of the image and one half of the distance from the bottom of the image.
  • [0038] 316 Extract a 3D connected region. Using the seed as a starting point, mark all values that are connected to the seed, that have the same value as the seed. Other selected connectivity algorithms are also suitable. An exemplary technique is disclosed in U.S. Pat. No. 4,751,643—METHOD AND APPARATUS FOR DETERMINING CONNECTED SUBSTRUCTURES WITHIN A BODY.
  • [0039] 317 Extract surfaces. Extract a surface comprised of triangles using an isosurface extraction technique. The isosurface corresponds to a value midway between the foreground and background values. Any isosurface extraction technique can be used. An exemplary isosurface extraction technique is the well-known Marching Cubes algorithm as described in U.S. Pat. No. 4,710,876—SYSTEM AND METHOD FOR THE DISPLAY OF SURFACE STRUCTURES CONTAINED WITHIN THE INTERIOR REGION OF A SOLID BODY.
  • The 2D contour identification proceeds as follows. Steps [0040] 310-316 correspond to steps 310-316 for the 3D surface.
  • [0041] 310 Acquire a CT data set that cover the lungs.
  • [0042] 311 Read the CT data set into memory.
  • [0043] 312 Select a threshold.
  • [0044] 313 Segment the study into foreground and background regions.
  • [0045] 314 Remove islands in the xy, xz and yz planes.
  • [0046] 315 Select a seed for the pleural space.
  • [0047] 316 Extract a 3D connected region.
  • [0048] 317 Extract a clipped portion of the volume data set that corresponds to the right pleural space. The clipped region should extent beyond the centerline of the data by a fixed percentage. This is to accommodate plural cavities that may cross the centerline. A 20% overlap seems appropriate for lung studies.
  • [0049] 318 Identify the contours in the left plural space. Extract contours comprised of line segments using a contour extraction technique. Any contour extraction technique can be used. An exemplary embodiment is the Marching Squares algorithm, a specialization of Marching Cubes described in U.S. Pat. No. 4,710,876—SYSTEM AND METHOD FOR THE DISPLAY OF SURFACE STRUCTURES.
  • [0050] 319 Sort the contours by line segment count and keep the contour with the largest number of line segments. This contour corresponds to the right pleural space.
  • [0051] 320 Extract a clipped portion of the volume data set that corresponds to the left pleural space. This is the same as step 318 except the clipping region is specified from the right side of the images.
  • [0052] 321 Identify the contours of the left plural space. This is the same as step 318, applied to the region in step 320. This contour corresponds to the left pleural space.
  • Employing the above described embodiments of the segmentation process, enables an automatic selection of all algorithm parameters based on the specific anatomy of the lung and the CT examination protocol. Further, the island removal is performed in three consecutive second passes, each in a different plane. It is to be appreciated that identifying the lung region initially allows a reduction in computation time and complexity for the downstream measurements. [0053]
  • Referring to FIG. 4, an embodiment of a hierarchy of models to be used in the method of FIG. 2 and a method for processing within the hierarchy are provided. The hierarchy of models comprise models of various levels comprising signal model data, geometric model data, and anatomical model data. Those pixels that have been classified as being inside the lung region at step [0054] 220 (FIG. 1) are modeled at several levels of modeling structure, herein after referred to as the hierarchy. As used herein, models refer generally to mathematical representations or, alternatively, mathematical translations.
  • At a first or low level, characteristics of the imaging device are translated into mathematical representations. Characteristics of the imaging device that are of interest are those characteristics that generally affect the display and resolution of the images or otherwise affect a radiologist's interpretation of regions in the image. For example, the scanner point spread function is a measurable indicator of the image formation process and may be mathematically modeled. Other indicators of the image formation process include X-ray density, brightness, resolution and contrast. [0055]
  • At a second or intermediate level, fitted shape models are derived to explain the geometry and intensity surface of various tissues. Shape and geometric model information is derived from anatomical information and expert radiologist observations which will be described in greater detail with respect to FIG. 4. [0056]
  • With one pass through the hierarchy, low-level pixel information (X-ray density) is transformed into anatomical information. This anatomical information is a classification of all pixels into lung tissue types, e.g. blood vessel, lung matrix, and lung cancer nodule. The models for the intermediate level are generally derived from pathological information for a region of interest (e.g. lung) and a specific disease (e.g. lung cancer or COPD) obtained from expert information, for example radiologists that have observed recurring characteristics for certain types of lung disease. The expert information is desirably from a radiologist or a plurality of radiologists experienced with the region of interest and detection of the specific disease. [0057]
  • The models described above create a hierarchy of information objects of increasing radiological importance that are explicitly modeled. For example, as is shown in FIG. 4, lung nodules and vascular structure are indicators of lung disease such as lung cancer. Additionally, lung parenchyma metrics are also indicators of lung disease. Based on radiologists” observations, or alternatively, other disease experts, a set of geometric and shape characteristics are obtained. For example, lung cancer nodules are generally compact, bright and spherical in nature. Further, lung cancer nodules that are likely to be cancerous tend to be spiculated (spidery vessel structures). In embodiments of the present invention, these characterizations of disease, such as lung cancer nodules, are mathematically represented as nodule model [0058] 470 and vessel model 480 as shown in FIG. 4. Of further interest is lung matrix tissue which can be considered background to the vessels and nodules and for embodiments of the present invention is also modeled as the mathematical representation of lung matrix tissue model 490.
  • Referring further to FIG. 4, nodule model [0059] 470, vessel model 480 and lung matrix tissue model 490 represent a high level explanation in the hierarchy used to distinguish various lung tissues. Each of the high level models are further defined at low and intermediate levels. For example, nodules are generally spherical and bright (measurable in Hounsfeld units). Thus, a shape model representing intensity region formation 440 and a signal model for step edge detection 410 are derived mathematically to enable identification of a potential nodule. In a further example, spiculated nodules tend to have a compact core structure fed by one or more vessels and also having a spidery or spiculated structure. A shape model representing intensity ribbon formation 450 and a signal model representing fold edge detection 420 similarly enable identification of a potential spiculated nodule. Background lung tissue is also similarly defined by low and intermediate levels of a texture region formation model 460 and a sub-resolution texture detection model 430.
  • At each level of the hierarchy, decisions are made via estimation of model parameters as to the characteristics of each information object, that is, each increasingly high-level explanation of the pixels in the images. The modeling steps at each level of the hierarchy. [0060]
  • 1) Pixel information as the output of an image formation process; and, [0061]
  • 2) Fitted shape models are derived to explain the geometry and intensity surface of various tissues. [0062]
  • In the first modeling step (step edge detection [0063] 410), pixel information is analyzed at the output of an image formation process (240 of FIGS. 1 and 2). The tissue boundaries are identified using convolution operators. Nodule candidates are localized by convolving the images with differential kernels defined by the signal impulse response of the imaging device. In an embodiment, images acquired from a GE LightSpeed Scanner were used and a Canny edge detector was used with a smoothing parameter of 1.1 pixels. An embodiment of the Canny edge-detection algorithm can be found in a description of Canny86.
  • The vascular structure is localized by convolving the images with differential kernels defined by the signal impulse response of the imaging device using fold edge detection [0064] 420. In this embodiment using images acquired from a GE LightSpeed Scanner, a fold edge detector was used with a smoothing parameter of 1.5 pixels.
  • Background tissue is represented as sub-resolution texture by sub-resolution texture detection [0065] 430. Background tissue is localized by identifying regions of low intensity. Convolution kernels defined by the signal impulse response of the imaging device are used to identify potential background regions. In this embodiment using images acquired from a GE LightSpeed Scanner, a Canny edge detector is used with a smoothing parameter of 1.1 pixels. At this stage of processing, the list of background regions is trimmed by thresholding at an average intensity of 520. An alternate localization procedure consists of modeling the background tissue as generalized intensity cylinders with random orientation. In this implementation, localization is achieved by comparing the output of generalized-cylinder model with the image intensities.
  • In the second modeling step, fitted shape models are used to explain the geometry and intensity surface of the various tissues. Putative nodule candidates are formed by grouping the output of the signal model stage into regions at intensity region formation step [0066] 440. Region grouping is performed by extrapolating edge segments perpendicular to the edge gradient. Edges ending near vertices associated with other edges are connected to form regions. In this implementation, the distance threshold for connecting edge segments is 4 pixels.
  • The vascular structure is obtained at intensity ribbon formation step [0067] 450 by linking together the output of the fold edge detection 420. At each point on the chain, a width of the chain is defined by locating the nearest step-edge on each side in a direction perpendicular to the chain direction. Through this sweeping operation, a set of intensity ribbons is defined. These ribbons are implicitly defined by the centerline of the fold-edges and the width of the fold along its entire length. These ribbons are considered “candidate vessels”, that is, objects which may be defined as blood vessels in the next level of the hierarchy.
  • At step [0068] 460 (Texture region formation), background lung tissue and lung matrix tissue are modeled. Background lung tissue is obtained by grouping together regions output by the signal operators. Regions are formed by extrapolating edge segments perpendicular to the edge gradient. Edges ending near vertices associated with other edges are connected to form regions. In this implementation, the distance threshold for connecting edge segments is 4 pixels.
  • Turning now to the decision process to determine which of the candidate nodules found in modeling step [0069] 2 are true lung cancer nodules. At this point, there are essentially two competing segmentations of the CT image—a region segmentation and a ribbon segmentation. Each region is a candidate nodule, and it must be decided whether the region, with an appropriate model on pixel intensities and region shape, is a better explanation of its interior pixels than any possible vessel or background explanation. To accomplish this, the two models are compared at step 500 using the Bayes Factor. The competition framework is a pair wise comparison of the modeled information: nodule vs. vessel, and nodule vs. background. If the nodule “wins” each competition, then it is considered a suspicious region and is reported as such.
  • As used herein, “Bayes Factors” refer to a known decision mechanism to ensure that the optimal decision is made given the input parameters. Applying the Bayes Factor to embodiments of the present invention provides that optimal decisions will be made given the statistical models of the shapes and signals provided by the radiologists' expert observations. This optimality assumes the statistical models of each anatomy type represent all the relevant knowledge embodied in a trained radiologist, and that the radiologist acts in a rational manner as defined by the Bayes Factor mechanism. Thus, the hierarchy of information enables processing to make a same decision as a radiologist would make regarding a region or nodule. Also, as used herein Bayes Factors will be used interchangeably herein with the term “Bayesian model competition”. [0070]
  • To begin the competition, for each candidate nodule a patch of pixels around the candidate is considered. In the nodule vs. vessel competition, this patch is defined as the union of the candidate nodule and each conflicting ribbon, in turn. In the nodule vs. background competition, the patch is defined as all pixels below a pre-specified intensity threshold within a pre-specified radius of the geometric center of the candidate, unioned with the candidate nodule. The radius is desirably set to 10 pixels, and the intensity threshold is desirably set to 520 CT units. Once the competition patch has been defined, each of the two competitions proceed in the same fashion. The following table gives all the information that is available during processing. [0071]
    TABLE 1
    Notation Description
    x intensity data
    θ1 geometrical measurement on “model 1”, eg. “one
    nodule, background elsewhere”
    θ2 geometrical measurement on model 2, eg. “one
    ribbon, background elsewhere”
    p(x|M = 1, β1, θ1) intensity model given that Model 1 is correct (β is
    a nuisance parameter)
    p(x|M = 2, β2, θ2) intensity model given that Model 2 is correct
    p(β1|M = 1) prior distribution on the intensity model nuisance
    parameter, given model 1 is correct
    p(β2|M = 2) prior distribution on the intensity model nuisance
    parameter, given model 2 is correct
    p(θ1|M = 1) prior distribution on the geometrical parameters in
    model 1
    p(θ2|M = 2) prior distribution on the geometrical parameters in
    model 2
    p(M = 1) prior probabilities on model 1 (the “incidence” of
    model 1)
    p(M = 2) prior probabilities on model 2
  • In each row of Table 1, the notation p(a|b) indicates the conditional probability distribution on the random variable a given the value of the random variable b. [0072]
  • The intensity model x[0073] i, given the candidate nodule is correct as a normal distribution with mean equal to a 2 parameter parabaloid, and constant variance, is defined as:
  • x i01 ui*22 v i *2i  (1)
  •  where (u[0074] i,vi) is the two-dimensional location of the ith pixel, after a rotation and translation that forces the least squares estimates of all other parameters in the full 6-parameter parabaloid in two dimensions to be zero. The error term ε1 is normally distributed with zero mean and fixed variance, estimated off-line from true nodule data. As used herein, “off-line” estimation refers to known information learned or known beforehand, such as chances or likelihood information.
  • The intensity model given the vessel is correct as a two-parameter parabaloid is defined as: [0075]
  • xi01 u i *2i  (2)
  •  where in this case u[0076] i is defined as a unit vector in the direction normal to the fold-edge chain direction. The fold edge center of the ribbon is defined as ui=0. Spans are defined at one-pixel separation along the chain, and each span's intensity data is modeled independently according to the above model. The error term is again normally distributed.
  • The background model is defined as independent normal data at each pixel with unknown mean and fixed variance, estimated off-line from true background data. This data is gathered by an expert and the variance is estimated using the usual normal-model unbiased estimate. [0077]
  • Prior distributions are defined on all intensity model parameters as normal distributions with means and covariance matrices estimated off-line from manually segmented intensity data. Prior distributions on shape parameters are defined as uniform distributions on key shape characteristics like nodule aspect ratio and size. Prior probabilities on each model are determined via a known scanner parameter known as Receiver Operating Characteristic curves according to pre-specified sensitivity and specificity targets. [0078]
  • To decide the winner of each competition, the Bayes Factor is calculated. The Bayes Factor as used herein refers to the ratio of posterior model probabilities given the intensity and shape data calculated in the last level (Step [0079] 3) of modeling hierarchy for two given models M=1 and M=2, p ( M = 1 x , θ 1 , θ 2 ) p ( M = 2 x , θ 1 , θ 2 ) . ( 3 )
    Figure US20030095692A1-20030522-M00001
  • This ratio can be written as [0080] p ( M = 1 | x , θ 1 , θ 2 ) p ( M = 2 x , θ 1 , θ 2 ) = p ( x | θ 1 , M = 1 ) p ( θ 1 | M = 1 ) p ( θ 2 | θ 1 , M = 1 ) p ( M = 1 ) p ( x | θ 2 , M = 2 ) p ( θ 2 | M = 2 ) p ( θ 1 | θ 2 , M = 2 ) p ( M = 2 ) ( 4 )
    Figure US20030095692A1-20030522-M00002
  • where the factor [0081] p ( θ 2 | θ 1 , M = 1 ) p ( θ 1 | θ 2 , M = 2 ) ( 5 )
    Figure US20030095692A1-20030522-M00003
  •  is assumed equal to one. The Bayes factor is necessarily greater than zero, and it indicates evidence for model [0082] 1 if the factor is greater than one (and vice versa).
  • Candidate nodules which give Bayes factors greater than one in both competitions are deemed suspicious, and are superimposed the CT data in a visualization tool (presenting step [0083] 280 of FIG. 2). The characteristics of these suspicious nodules are also stored for further follow-up.
  • It is to be appreciated that incorporating knowledge about the imaging processing and the imaging device into the analysis techniques increases the accuracy and robustness of image measurements. The competition framework provides a robust method for making a model selection decision. Modeling the anatomy in the images improves the robustness of the image measurements and allows results to be presented to doctors in the context of the anatomy. The anatomical models are easily explained to physicians and their expert knowledge is coherently incorporated into the system (in the form of mathematical approximations of anatomical features). The lowest level of the modeling hierarchy relies on time-tested image formation and understanding techniques which are firmly grounded in human visual perception. Anatomical models are chosen via Bayes Factors, enable optimal decision given our statistical models. No training data (such as the voluminous training data needed for neural networks) is required, but it can be used to supplement expert knowledge in the specification of prior distributions. The model-based approach allows incorporation of expert information elicited from radiologists. The results are reported to a radiologist or doctor incorporating the anatomical context of each suspicious region and the decision process by which suspicious regions were identified. [0084]
  • In a further embodiment, the results are reported to a radiologist or doctor (hereinafter “user”) in a manner that the user receives anatomical context, reasons for the decision of whether the region is of a particular type (nodule or vessel), and information of importance to the user. Information of radiological importance are, for example, size of nodule, number of vessels, evidence of spiculation, chances/likelihood of cancer or disease, brightness measurements and other characteristic information related to the disease at issue. Processor [0085] 120 of FIG. 1 is adapted to perform the computations needed to support this reporting functionality. In a further embodiment, processor 120 is adapted to allow for user queries regarding particular regions of interest such as pointing to a region and receiving information such as size, number of vessels and brightness for the selected region.
  • The embodiments above related to lung cancer detection and specifically to the distinction between lung nodules and vessels. In further embodiments, additional lung disease characteristics are similarly modeled such as the low-density, sponge-like texture which is generally characteristic of emphysema. As was described with reference to nodules and vessels, anatomical feature descriptions are obtained by experts (e.g. radiologists) and mathematically represented as a hierarchy. In further embodiments, models are derived for diseases that occur in other areas, such as the brain, colon and heart. [0086]
  • Also, in a further embodiment, the hierarchy of models may be used in known neural network techniques as the training data to identify low and intermediate information and prior distributions. It is desirable that the Bayes Factor analysis be applied at the higher levels to provide useful and interpretative diagnosis data and the decision process. [0087]
  • In a further embodiment, processor [0088] 120 is further adapted to store the anatomical context and processed image data sets to be searched and retrieved remotely. In this embodiment, the information developed at each level in the model hierarchy is stored in systems used for medical archives, medical search and retrieval systems and alternate medical disease reporting systems. Further in this embodiment, information that may be searched and retrieved include: pathological and anatomical models derived for characteristics of diseases, images representative of the diseases, and results of the model hierarchy computations (processed image data sets). The capability of storing/retrieving information for a particular diseased tissue type enables broader access to the information, such as via the Internet, a hospital information system, a radiological information system, or other information transmission infrastructure. Additionally, this information allows matching and retrieval of exams classified as similar based on the information provided by model hierarchy computations.
  • In further embodiments, processor [0089] 120 is adapted to automatically send detailed exam information to remote workstations or portable computing device via an information transmission infrastructure. In a further embodiment of processor 120, processor 120 is adapted to automatically send detailed exam information which meets selected specified requirements determined in advance of transmission or determined adaptively by the processing system. In order to further tune or adjust analysis programs, processor 120 is also adapted to tune at least one computer analysis algorithm based on information from model hierarchy computations stored in previous exams.
  • Also, in another further embodiment, processor [0090] 120 is further adapted to generate statistical measurements based on the information from model hierarchy computations stored in previous exams and report results of the statistical measurements to a local or remote monitoring facility. In this embodiment, processor 120 may also be configured to report the results of the statistical measurements if predetermined criteria based on the system performance are met.
  • In an exemplary embodiment, the steps outlined in the last section are implemented in C++ code based on the Targetjr image understanding library (http://www.targetjr.org). A set of DICOM (Digital Image and Communication in Medicine) image files, one for each slice in the CT scan, are input into the program and the program returns suspicious nodules to be visualized on the original CT data or saved for further follow-up. It is to be appreciated that other coding software known to one skilled in the art would not depart from the spirit of the invention. [0091]
  • The embodiments of the invention presented in previous paragraphs focus on the problem of locating suspicious regions in CT lung scans. It is to be appreciated that the hierarchical image modeling framework can be directly transferred to other imaging modalities (for example MRI, X-ray, ultrasound scanner, positron emission tomography (PET) scanner) and diseases by re-specifying the low-level detection techniques and the statistical distributions of anatomy. [0092]
  • While the preferred embodiments of the present invention have been shown and described herein, it will be obvious that such embodiments are provided by way of example only. Numerous variations, changes and substitutions will occur to those of skill in the art without departing from the invention herein. Accordingly, it is intended that the invention be limited only by the spirit and scope of the appended claims. [0093]

Claims (58)

  1. 1. A method for processing medical images for use in the detection and diagnosis of disease comprising:
    classifying regions of interest within the medical images based on a hierarchy of anatomical models representative of anatomical information indicative of a given disease and signal models of signal information of an image acquisition device used to acquire the medical images.
  2. 2. The method of claim 1 further comprising the step of presenting the regions of interest from the classifying step, wherein the presenting step comprises at least one of displaying the regions of interest, displaying a plurality of decision parameters and displaying anatomical context of the regions of interest.
  3. 3. The method of claim 1 wherein the hierarchy of anatomical models comprise at least one geometric information, shape information, pixel intensity, size information and anatomical information derived from expert observation and knowledge of the given disease and further derived from expert description of characteristics of the given disease.
  4. 4. The method of claim 1 wherein the signal models comprises at least one of a scanner point spread function or the impulse response of the image acquisition device and the X-ray density, brightness, resolution and contrast of anatomical structures.
  5. 5. The method of claim 1 wherein the classifying step further comprises the step of comparing respective anatomical models using Bayes Factors to differentiate tissue types in the images.
  6. 6. The method of claim 1 wherein the classifying step further comprises image segmentation processing.
  7. 7. The method of claim 5 further comprises presenting results of tissues types differentiated in the comparing step.
  8. 8. The method of claim 1 wherein the image acquisition device is selected from at least one of a computed tomography (CT) scanner, a magnetic resonance imaging (MRI) scanner, ultrasound scanner, positron emission tomography scanner and a X-ray device.
  9. 9. The method of claim 1 wherein the anatomical models correspond to a characteristics of diseased lung tissue.
  10. 10. A method for identification of at least one suspect region of interest in acquired medical images of a subject, comprising:
    grouping a plurality of mathematical models corresponding to anatomical and tissue characteristics into a hierarchy of anatomical models for tissue and region differentiation, the anatomical models being selected to be representative of anatomical information indicative of a given disease;
    classifying regions of interest within the medical images based on the hierarchy of anatomical models and further based on signal models corresponding to a given image acquisition device used to acquire the medical images; and, presenting the results of the competing step for use in identifying suspicious regions in the medical images.
  11. 11. The method of claim 10 wherein the signal models comprises at least one of a scanner point spread function or the impulse response of the image acquisition device and the X-ray density, brightness, resolution and contrast of anatomical structures.
  12. 12. The method of claim 10 wherein the classifying step further comprises the step of comparing respective anatomical models in pairs using Bayes Factors to differentiate tissue types in the images.
  13. 13. The method of claim 10 wherein the classifying step is performed by a combination of Bayesian competition and neural networks.
  14. 14. The method of claim 10 wherein the presenting step comprises at least one of displaying the regions of interest, displaying a plurality of decision parameters and displaying anatomical context of the suspicious regions in the medical images.
  15. 15. The method of claim 10 further comprising;
    interfacing with a user during the presenting step wherein the user inputs queries regarding the anatomical context of each suspicious region and the decision process.
  16. 16. The method of claim 10 wherein the hierarchy of anatomical models comprise at least one geometric information, shape information, pixel intensity, size information and anatomical information derived from expert observation and knowledge of the given disease and further derived from expert description of characteristics of the given disease.
  17. 17. The method of claim 16 wherein the hierarchy of anatomical models are derived for lung nodules, the vascular structure, and lung parenchyma.
  18. 18. The method of claim 10 wherein the plurality of medical images are acquired by an image acquisition device and the image acquisition device is selected from at least one of a computed tomography (CT) scanner, a magnetic resonance imaging (MRI) scanner, a X-ray device, an ultrasound scanner, and a positron emission tomography (PET) scanner.
  19. 19. The method of claim 10 wherein the medical images are of a lung.
  20. 20. A method for identifying and quantifying tissue types from a plurality of acquired medical images for use in diagnosing lung disease comprising
    segmenting the acquired image data to define the lung region;
    computing low level features in the images using the known characteristics of the imaging device and the imaging process;
    computing a hierarchy of anatomical models representative of anatomical features in the images;
    grouping regions in the image into anatomical structures; and,
    identifying areas within the grouped regions representing areas which are suspicious for at least one given lung disease.
  21. 21. The method of claim 20 further comprising presenting the areas identified as suspicious for lung disease.
  22. 22. The method of claim 21 wherein the presenting step comprises presenting an anatomical context of an identified suspicious area and a decision process by which the suspicious area was identified.
  23. 23. The method of claim 20 wherein the medical images are acquired from at least one of a computed tomography (CT) scanner, a magnetic resonance imaging (MRI) scanner, an ultrasound scanner, and a positron emission tomography scanner, or a X-ray device.
  24. 24. The method of claim 20 wherein the grouping of regions into anatomical structures is performed by Bayesian competition in accordance with the hierarchy of anatomical models.
  25. 25. A method for characterizing tissue in medical images for use in disease diagnosis and detection comprising:
    computing an information object hierarchy of increasing complexity to characterize anatomical tissue, the object hierarchy containing models based on characteristics an image acquisition device and anatomical characteristics of at least one given disease;
    comparing information using Bayes Factors at levels of the object hierarchy to identify suspicious tissue indicative of disease.
  26. 26. The method of claim 25 wherein the anatomical characteristics are at least one of geometric shape and intensity values.
  27. 27. The method of claim 25 wherein the object hierarchy further comprises a plurality of signal models based on characteristics of an image acquisition device used in acquiring the images and plurality of models based on anatomical characteristics of a selected region of interest and the at least one given disease.
  28. 28. The method of claim 25 wherein the hierarchy comprises a low level corresponding to signal models representative of the image acquisition device and a high level corresponding to anatomical models derived in accordance with expert observation and knowledge of the at least one given disease.
  29. 29. The method of claim 25 further comprising a plurality of intermediate levels defining geometric models, shape models, intensity models derived from the anatomical models at the high level.
  30. 30. The method of claim 25 where the at least one given disease is lung disease.
  31. 31. The method of claim 28 wherein the anatomical models at the high level correspond to lung nodules indicative of lung disease, vascular structure and lung matrix tissue.
  32. 32. The method of claim 31 further comprising a plurality of intermediate levels defining shape models and intensity models characteristic of lung nodules.
  33. 33. The method of claim 31 further comprising a plurality of intermediate levels defining shape models and intensity models characteristic of lung vascular structure.
  34. 34. The method of claim 31 wherein the Bayes Factor is a ratio of posterior model probabilities given intensity and shape data for two given models M=1 and M=2, and where x=intensity data, θ1=geometric model for M=1 and θ2=geometric model for M=2 and is expressed as:
    p ( M = 1 | x , θ 1 , θ 2 ) p ( M = 2 | x , θ 1 , θ 2 ) .
    Figure US20030095692A1-20030522-M00004
  35. 35. A system for processing medical images acquired by an image acquisition device for use in the detection and diagnosis of disease comprising:
    a processor coupled to the image acquisition device, the processor is adapted to identify suspicious regions within the medical images based on an information object hierarchy and a Bayes Factors competition framework using at least one of anatomical models and signal models; and,
    an interface coupled to the processor adapted to present information relating to the suspicious regions identified by the processor, the information being used for diagnosis and detection.
  36. 36. The system of claim 35 wherein the anatomical models comprise at least one of lung nodules and vascular structure indicative of lung disease.
  37. 37. The system of claim 35 wherein the hierarchy comprises models of increasing complexity for use in identifying suspicious regions.
  38. 38. The system of claim 35 wherein the interface is further adapted to receive user queries regarding an anatomical context of the suspicious regions and a decision process for identifying each of the suspicious regions.
  39. 39. The system of claim 35 wherein the signal models comprises at least one of a scanner point spread function, the impulse response of the image acquisition device and the X-ray density, brightness, resolution and contrast of anatomical structures.
  40. 40. The system of claim 35 wherein the image acquisition device is selected from at least one of a computed tomography (CT) scanner, a magnetic resonance imaging (MRI) scanner, an ultrasound scanner, and a positron emission tomography scanner, and a X-ray device.
  41. 41. The system of claim 35 wherein the Bayes Factor competition uses the Bayes Factor, the Bayes Factor being to a ratio of posterior model probabilities given intensity and shape data for two given models M=1 and M=2, and where x=intensity data, θ1=geometric model for M=1 and θ2=geometric model for M=2 and is expressed as:
    p ( M = 1 | x , θ 1 , θ 2 ) p ( M = 2 | x , θ 1 , θ 2 ) .
    Figure US20030095692A1-20030522-M00005
  42. 42. The system of claim 35 wherein the processor is further adapted to automatically segment the pleural space.
  43. 43. The system of claim 35 wherein the processor is further adapted to group a plurality of anatomical and signal models into the hierarchy of models for use in the competition framework.
  44. 44. The system of claim 35 wherein the medical images are acquired of at least one of a lung, a colon, a breast, a brain and a limb.
  45. 45. A computer-aided system for use in the diagnosis and detection of disease comprises:
    an image acquisition device for acquiring a plurality of image data sets; and,
    a processor adapted to process the image data sets and to classify selected tissue types within the image data sets based on a hierarchy of signal and anatomical models and the processor is further adapted to differentiate anatomical context of the classified tissue types for use in the diagnosis and detection of disease.
  46. 46. The system of claim 45 further comprising an interface unit for presenting the classified tissue types within the image data sets and the anatomical context of the classified tissue types for aiding an interpretation of the processed image data sets.
  47. 47. The system of claim 45 wherein the image acquisition device is at least one of a computed tomography (CT) X-ray system, a magnetic resonance imaging (MRI) system, an ultrasound scanner, and a positron emission tomography scanner, and a X-ray system.
  48. 48. The system of claim 45 wherein the anatomical context comprises at least one of lung nodules and vascular structures indicative of lung cancer, healthy lung tissue, and diseased lung tissue indicative of chronic obstructive pulmonary disease (COPD).
  49. 49. The system of claim 45 wherein the hierarchy comprises models of increasing complexity for use in identifying regions indicative of a given disease.
  50. 50. The system of claim 46 wherein the interface unit is further adapted to receive user queries regarding anatomical context indicative of a suspicious region and a decision process for identifying the suspicious region.
  51. 51. The system of claim 45 wherein the image data sets are acquired of at least one of a lung, a colon, a breast, a brain and a limb.
  52. 52. The system of claim 45 wherein the processor is further adapted to store the anatomical context and processed image data sets to be searched and retrieved via at least one of the Internet, a hospital information system, a radiological information system, and other information transmission infrastructure.
  53. 53. The system of claim 45 wherein the processor is further adapted to automatically send detailed exam information to remote workstations or portable computing device via an information transmission infrastructure.
  54. 54. The system of claim 53 wherein the processor is further adapted to automatically send detailed exam information which meets selected specified requirements determined in advance of transmission or determined adaptively by the processing system.
  55. 55. The system of claim 45 wherein the processor is further adapted to tune at least one computer analysis algorithm based on information from model hierarchy computations stored in previous exams.
  56. 56. The system of claim 45 wherein the processor is further adapted to generate statistical measurements based on the information from model hierarchy computations stored in previous exams.
  57. 57. The system of claim 56 wherein the processor is further adapted to report results of the statistical measurements to a local or remote monitoring facility.
  58. 58. The system of claim 57 wherein the processor is further adapted to report the results of the statistical measurements if predetermined criteria based on the system performance are met.
US09683111 2001-11-20 2001-11-20 Method and system for lung disease detection Expired - Fee Related US7058210B2 (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US09683111 US7058210B2 (en) 2001-11-20 2001-11-20 Method and system for lung disease detection

Applications Claiming Priority (4)

Application Number Priority Date Filing Date Title
US09683111 US7058210B2 (en) 2001-11-20 2001-11-20 Method and system for lung disease detection
DE2002626942 DE60226942D1 (en) 2001-11-20 2002-11-14 Image processing method and system for detecting diseases
EP20020257868 EP1315125B1 (en) 2001-11-20 2002-11-14 Image processing method and system for disease detection
JP2002335867A JP4310099B2 (en) 2001-11-20 2002-11-20 Method and system for lung disease detection

Publications (2)

Publication Number Publication Date
US20030095692A1 true true US20030095692A1 (en) 2003-05-22
US7058210B2 US7058210B2 (en) 2006-06-06

Family

ID=24742614

Family Applications (1)

Application Number Title Priority Date Filing Date
US09683111 Expired - Fee Related US7058210B2 (en) 2001-11-20 2001-11-20 Method and system for lung disease detection

Country Status (4)

Country Link
US (1) US7058210B2 (en)
EP (1) EP1315125B1 (en)
JP (1) JP4310099B2 (en)
DE (1) DE60226942D1 (en)

Cited By (43)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20030103664A1 (en) * 2001-12-05 2003-06-05 Guo-Qing Wei Vessel-feeding pulmonary nodule detection by volume projection analysis
US20030190064A1 (en) * 2002-04-03 2003-10-09 Hitoshi Inoue Radiographic image processing method, radiographic image processing apparatus, radiographic image processing system, program, computer-readable storage medium, image diagnosis assisting method, and image diagnosis assisting system
US20040109594A1 (en) * 2002-12-10 2004-06-10 Eastman Kodak Company Method for automatic construction of 2D statistical shape model for the lung regions
US20040109595A1 (en) * 2002-12-10 2004-06-10 Eastman Kodak Company Method for automated analysis of digital chest radiographs
US20050105788A1 (en) * 2003-11-19 2005-05-19 Matthew William Turek Methods and apparatus for processing image data to aid in detecting disease
US20050165294A1 (en) * 2003-03-18 2005-07-28 Weiss Kenneth L. Automated brain MRI and CT prescriptions in Talairach space
US20050169536A1 (en) * 2004-01-30 2005-08-04 Vittorio Accomazzi System and method for applying active appearance models to image analysis
US20050209888A1 (en) * 2004-03-09 2005-09-22 Naoki Oowaki X-ray exposure report system, medical apparatus, and examination protocol distribution system
US20060083415A1 (en) * 2004-10-15 2006-04-20 Georges El Fakhri Factor analysis in medical imaging
US20070014472A1 (en) * 2005-07-18 2007-01-18 Zhengrong Ying Method of and system for classifying objects using local distributions of multi-energy computed tomography images
US20070081700A1 (en) * 2005-09-29 2007-04-12 General Electric Company Systems, methods and apparatus for creation of a database of images from categorical indices
US20070092864A1 (en) * 2005-09-30 2007-04-26 The University Of Iowa Research Foundation Treatment planning methods, devices and systems
US20070127802A1 (en) * 2005-12-05 2007-06-07 Siemens Corporate Research, Inc. Method and System for Automatic Lung Segmentation
US20070242863A1 (en) * 2006-04-13 2007-10-18 Bernice Eland Hoppel Methods and Apparatus for Contouring at Least One Vessel
US20070246047A1 (en) * 2006-04-19 2007-10-25 Draeger Medical Ag & Co. Kg Process and device for lung ventilation
US20070258630A1 (en) * 2006-05-03 2007-11-08 Tobin Kenneth W Method and system for the diagnosis of disease using retinal image content and an archive of diagnosed human patient data
US20080062262A1 (en) * 2005-05-11 2008-03-13 Luc Perron Apparatus, method and system for screening receptacles and persons
US20080205718A1 (en) * 2005-05-23 2008-08-28 Koninklijke Philips Electronics, N.V. Automated Organ Linking for Organ Model Placement
US20090216794A1 (en) * 2008-02-27 2009-08-27 General Electric Company Method and system for accessing a group of objects in an electronic document
US7734102B2 (en) 2005-05-11 2010-06-08 Optosecurity Inc. Method and system for screening cargo containers
US20100228143A1 (en) * 2009-03-09 2010-09-09 Drager Medical Ag & Co. Kg Apparatus and method to determine functional lung characteristics
US7899232B2 (en) 2006-05-11 2011-03-01 Optosecurity Inc. Method and apparatus for providing threat image projection (TIP) in a luggage screening system, and luggage screening system implementing same
US7929737B2 (en) 2005-09-29 2011-04-19 General Electric Company Method and system for automatically generating a disease severity index
US7991242B2 (en) 2005-05-11 2011-08-02 Optosecurity Inc. Apparatus, method and system for screening receptacles and persons, having image distortion correction functionality
US20110222747A1 (en) * 2008-05-14 2011-09-15 Koninklijke Philips Electronics N.V. Image classification based on image segmentation
US20110255761A1 (en) * 2007-06-26 2011-10-20 University Of Rochester Method and system for detecting lung tumors and nodules
CN102243759A (en) * 2010-05-10 2011-11-16 东北大学 Three-dimensional lung vessel image segmentation method based on geometric deformation model
EP2485194A2 (en) 2011-02-07 2012-08-08 Fujifilm Corporation Image processing apparatus, method and program
WO2012109658A2 (en) * 2011-02-11 2012-08-16 Emory University Systems, methods and computer readable storage mediums storing instructions for segmentation of medical images
EP2490176A1 (en) 2011-02-15 2012-08-22 Fujifilm Corporation Surgery-assistance apparatus, method and program
US8494210B2 (en) 2007-03-30 2013-07-23 Optosecurity Inc. User interface for use in security screening providing image enhancement capabilities and apparatus for implementing same
US8594402B2 (en) 2010-03-05 2013-11-26 Fujifilm Corporation Image diagnosis support apparatus, method and program
WO2014015424A1 (en) * 2012-07-26 2014-01-30 International Business Machines Corporation Tumor classification based on an analysis of a related ultrasonic attenuation map
US20140198892A1 (en) * 2011-03-22 2014-07-17 Keisuke Yamakawa X-ray ct device, and method
US8903147B2 (en) 2011-03-30 2014-12-02 Fujifilm Corporation Medical report generation apparatus, method and program
US8907949B2 (en) 2010-08-23 2014-12-09 Fujifilm Corporation Image display apparatus, method and program
US8953858B2 (en) 2010-01-28 2015-02-10 Radlogics, Inc. Methods and systems for analyzing, prioritizing, visualizing, and reporting medical images
CN104732086A (en) * 2015-03-23 2015-06-24 深圳市智影医疗科技有限公司 Computer-assisted disease detection system based on cloud computing
US9117291B2 (en) 2012-04-19 2015-08-25 Fujifilm Corporation Image processing apparatus, image processing method, and non-transitory storage medium
US9204853B2 (en) * 2012-10-11 2015-12-08 Carestream Health, Inc. Method and system for quantitative imaging
CN105701799A (en) * 2015-12-31 2016-06-22 东软集团股份有限公司 Method and device for segmenting lung blood vessel in lung mask image
US20170035381A1 (en) * 2015-08-06 2017-02-09 Case Western Reserve University Characterizing disease and treatment response with quantitative vessel tortuosity radiomics
US9632206B2 (en) 2011-09-07 2017-04-25 Rapiscan Systems, Inc. X-ray inspection system that integrates manifest data with imaging/detection processing

Families Citing this family (48)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6941323B1 (en) 1999-08-09 2005-09-06 Almen Laboratories, Inc. System and method for image comparison and retrieval by enhancing, defining, and parameterizing objects in images
JP2004041694A (en) * 2002-05-13 2004-02-12 Fuji Photo Film Co Ltd Image generation device and program, image selecting device, image outputting device and image providing service system
US7116810B2 (en) * 2002-11-27 2006-10-03 General Electric Company Method and system for airway measurement
US7653227B2 (en) * 2004-02-09 2010-01-26 Siemens Medical Solutions Usa, Inc. Hierarchical modeling in medical abnormality detection
DE10357205A1 (en) * 2003-12-08 2005-07-14 Siemens Ag A method of generating images of an examination object Result
JP2008507372A (en) * 2004-07-26 2008-03-13 コーニンクレッカ フィリップス エレクトロニクス エヌ ヴィ System and method for suspicious object boundary determination Automated
US7590271B2 (en) * 2004-10-28 2009-09-15 Siemens Medical Solutions Usa, Inc. System and method for automatic detection and localization of 3D bumps in medical images
GB2420641B (en) * 2004-11-29 2008-06-04 Medicsight Plc Digital medical image analysis
WO2006058738A9 (en) * 2004-12-02 2007-07-26 Hoe Lieven Van Image evaluation system, methods and database
US8064663B2 (en) 2004-12-02 2011-11-22 Lieven Van Hoe Image evaluation system, methods and database
US20060217925A1 (en) * 2005-03-23 2006-09-28 Taron Maxime G Methods for entity identification
FR2886433B1 (en) * 2005-05-30 2007-09-07 Commissariat Energie Atomique Method of segmenting a sequence of three-dimensional images, in particular pharmaco-imaging.
JP5133505B2 (en) * 2005-06-24 2013-01-30 ジーイー・メディカル・システムズ・グローバル・テクノロジー・カンパニー・エルエルシー Image determining apparatus and x-ray ct device
US8050734B2 (en) * 2005-09-07 2011-11-01 General Electric Company Method and system for performing patient specific analysis of disease relevant changes of a disease in an anatomical structure
US20070081699A1 (en) * 2005-09-29 2007-04-12 General Electric Company Systems, methods and apparatus for diagnosis of disease from categorical indices
US20070081701A1 (en) * 2005-09-29 2007-04-12 General Electric Company Systems, methods and apparatus for tracking progression and tracking treatment of disease from categorical indices
US7864995B2 (en) * 2006-02-11 2011-01-04 General Electric Company Systems, methods and apparatus of handling structures in three-dimensional images
US7864994B2 (en) * 2006-02-11 2011-01-04 General Electric Company Systems, methods and apparatus of handling structures in three-dimensional images having multiple modalities and multiple phases
US8626263B2 (en) * 2006-04-13 2014-01-07 General Electric Company Methods and apparatus for relative perfusion and/or viability
US7672496B2 (en) * 2006-06-21 2010-03-02 Icad, Inc. Forming three dimensional objects using a decision rule in medical image data
EP2052363A2 (en) * 2006-08-09 2009-04-29 Philips Electronics N.V. A method, apparatus, graphical user interface, computer-readable medium, and use for quantification of a structure in an object of an image dataset
US7920729B2 (en) * 2006-08-10 2011-04-05 General Electric Co. Classification methods and apparatus
EP2070045B1 (en) * 2006-09-22 2018-06-27 Koninklijke Philips N.V. Advanced computer-aided diagnosis of lung nodules
US8923577B2 (en) * 2006-09-28 2014-12-30 General Electric Company Method and system for identifying regions in an image
KR20100016039A (en) * 2007-03-30 2010-02-12 코닌클리케 필립스 일렉트로닉스 엔.브이. Improved treatment plan evaluation in radiotherapy by stochastic analysis of delineation uncertainty
US8989468B2 (en) * 2007-05-25 2015-03-24 Definiens Ag Generating an anatomical model using a rule-based segmentation and classification process
KR100882275B1 (en) 2007-05-25 2009-02-06 전남대학교산학협력단 A disease grouping system through the face image
US8265367B2 (en) * 2007-06-04 2012-09-11 Siemens Computer Aided Diagnostics, Ltd. Identifying blood vessels in lung x-ray radiographs
US8150135B2 (en) * 2007-06-04 2012-04-03 Siemens Computer Aided Diagnosis Ltd. Identifying ribs in lung X-rays
US8126229B2 (en) * 2007-08-03 2012-02-28 Siemens Medical Solutions Usa, Inc. Reduction of lymph tissue false positives in pulmonary embolism detection
US20090082637A1 (en) * 2007-09-21 2009-03-26 Michael Galperin Multi-modality fusion classifier with integrated non-imaging factors
JP5661468B2 (en) * 2007-11-14 2015-01-28 コーニンクレッカ フィリップス エヌ ヴェ Computer-aided detection of disease (cad)
US20090136111A1 (en) * 2007-11-25 2009-05-28 General Electric Company System and method of diagnosing a medical condition
CN102272799B (en) * 2008-09-26 2015-07-22 皇家飞利浦电子股份有限公司 Anatomy-defined automated CPR generation
KR101050769B1 (en) * 2009-05-08 2011-07-21 가톨릭대학교 산학협력단 Medical image processing system and method
JP5308973B2 (en) 2009-09-16 2013-10-09 富士フイルム株式会社 Medical image information displaying device and method and program
JP2011092681A (en) 2009-09-30 2011-05-12 Fujifilm Corp Medical image processor, method, and program
CN101713776B (en) 2009-11-13 2013-04-03 长春迪瑞医疗科技股份有限公司 Neural network-based method for identifying and classifying visible components in urine
EP2538839B1 (en) * 2010-02-23 2015-07-08 Loma Linda University Method of analyzing a medical image
GB2478329B (en) * 2010-03-03 2015-03-04 Samsung Electronics Co Ltd Medical image processing
DE102012200225A1 (en) * 2012-01-10 2013-07-11 Siemens Aktiengesellschaft Method for processing e.g. three-dimensional image data of body of patient using medical apparatus for recognizing e.g. tumor in region of fatty tissue, involves processing image data of region based on determined property of region
JP5797124B2 (en) 2012-01-31 2015-10-21 富士フイルム株式会社 Surgery support apparatus, surgery support method and surgical support program
JP6008635B2 (en) 2012-07-24 2016-10-19 富士フイルム株式会社 Surgery support apparatus, method, and program
RU2015117639A (en) * 2012-10-09 2016-12-10 Конинклейке Филипс Н.В. Multistructural atlas and / or its use
EP2988659A4 (en) * 2013-04-23 2017-01-04 Univ Of Maine System Board Of Trustees Improved methods of cancer detection
KR101575620B1 (en) * 2014-02-21 2015-12-08 전북대학교산학협력단 Detection method and system for region of interest using normalized signal intensity of medical image
KR101576058B1 (en) 2014-03-26 2015-12-10 전북대학교산학협력단 Method of Region Of Interest Selection and System using characteristics of MRI/MRA
KR20160037023A (en) * 2014-09-26 2016-04-05 삼성전자주식회사 Apparatus and Method for supporting a Computer Aided Diagnosis

Citations (18)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4710876A (en) * 1985-06-05 1987-12-01 General Electric Company System and method for the display of surface structures contained within the interior region of a solid body
US4751643A (en) * 1986-08-04 1988-06-14 General Electric Company Method and apparatus for determining connected substructures within a body
US4907156A (en) * 1987-06-30 1990-03-06 University Of Chicago Method and system for enhancement and detection of abnormal anatomic regions in a digital image
US5235510A (en) * 1990-11-22 1993-08-10 Kabushiki Kaisha Toshiba Computer-aided diagnosis system for medical use
US5331550A (en) * 1991-03-05 1994-07-19 E. I. Du Pont De Nemours And Company Application of neural networks as an aid in medical diagnosis and general anomaly detection
US5359513A (en) * 1992-11-25 1994-10-25 Arch Development Corporation Method and system for detection of interval change in temporally sequential chest images
US5437279A (en) * 1992-07-02 1995-08-01 Board Of Regents, The University Of Texas System Method of predicting carcinomic metastases
US5692507A (en) * 1990-07-02 1997-12-02 Varian Associates, Inc. Computer tomography apparatus using image intensifier detector
US5779634A (en) * 1991-05-10 1998-07-14 Kabushiki Kaisha Toshiba Medical information processing system for supporting diagnosis
US5832103A (en) * 1993-11-29 1998-11-03 Arch Development Corporation Automated method and system for improved computerized detection and classification of massess in mammograms
US5881124A (en) * 1994-03-31 1999-03-09 Arch Development Corporation Automated method and system for the detection of lesions in medical computed tomographic scans
US5943435A (en) * 1997-10-07 1999-08-24 Eastman Kodak Company Body part recognition in radiographic images
US5987345A (en) * 1996-11-29 1999-11-16 Arch Development Corporation Method and system for displaying medical images
US5987094A (en) * 1996-10-30 1999-11-16 University Of South Florida Computer-assisted method and apparatus for the detection of lung nodules
US6004267A (en) * 1997-03-07 1999-12-21 University Of Florida Method for diagnosing and staging prostate cancer
US6125194A (en) * 1996-02-06 2000-09-26 Caelum Research Corporation Method and system for re-screening nodules in radiological images using multi-resolution processing, neural network, and image processing
US6272470B1 (en) * 1996-09-03 2001-08-07 Kabushiki Kaisha Toshiba Electronic clinical recording system
US6574304B1 (en) * 2002-09-13 2003-06-03 Ge Medical Systems Global Technology Company, Llc Computer aided acquisition of medical images

Patent Citations (18)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4710876A (en) * 1985-06-05 1987-12-01 General Electric Company System and method for the display of surface structures contained within the interior region of a solid body
US4751643A (en) * 1986-08-04 1988-06-14 General Electric Company Method and apparatus for determining connected substructures within a body
US4907156A (en) * 1987-06-30 1990-03-06 University Of Chicago Method and system for enhancement and detection of abnormal anatomic regions in a digital image
US5692507A (en) * 1990-07-02 1997-12-02 Varian Associates, Inc. Computer tomography apparatus using image intensifier detector
US5235510A (en) * 1990-11-22 1993-08-10 Kabushiki Kaisha Toshiba Computer-aided diagnosis system for medical use
US5331550A (en) * 1991-03-05 1994-07-19 E. I. Du Pont De Nemours And Company Application of neural networks as an aid in medical diagnosis and general anomaly detection
US5779634A (en) * 1991-05-10 1998-07-14 Kabushiki Kaisha Toshiba Medical information processing system for supporting diagnosis
US5437279A (en) * 1992-07-02 1995-08-01 Board Of Regents, The University Of Texas System Method of predicting carcinomic metastases
US5359513A (en) * 1992-11-25 1994-10-25 Arch Development Corporation Method and system for detection of interval change in temporally sequential chest images
US5832103A (en) * 1993-11-29 1998-11-03 Arch Development Corporation Automated method and system for improved computerized detection and classification of massess in mammograms
US5881124A (en) * 1994-03-31 1999-03-09 Arch Development Corporation Automated method and system for the detection of lesions in medical computed tomographic scans
US6125194A (en) * 1996-02-06 2000-09-26 Caelum Research Corporation Method and system for re-screening nodules in radiological images using multi-resolution processing, neural network, and image processing
US6272470B1 (en) * 1996-09-03 2001-08-07 Kabushiki Kaisha Toshiba Electronic clinical recording system
US5987094A (en) * 1996-10-30 1999-11-16 University Of South Florida Computer-assisted method and apparatus for the detection of lung nodules
US5987345A (en) * 1996-11-29 1999-11-16 Arch Development Corporation Method and system for displaying medical images
US6004267A (en) * 1997-03-07 1999-12-21 University Of Florida Method for diagnosing and staging prostate cancer
US5943435A (en) * 1997-10-07 1999-08-24 Eastman Kodak Company Body part recognition in radiographic images
US6574304B1 (en) * 2002-09-13 2003-06-03 Ge Medical Systems Global Technology Company, Llc Computer aided acquisition of medical images

Cited By (74)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US7020316B2 (en) * 2001-12-05 2006-03-28 Siemens Corporate Research, Inc. Vessel-feeding pulmonary nodule detection by volume projection analysis
US20030103664A1 (en) * 2001-12-05 2003-06-05 Guo-Qing Wei Vessel-feeding pulmonary nodule detection by volume projection analysis
US7158661B2 (en) * 2002-04-03 2007-01-02 Canon Kabushiki Kaisha Radiographic image processing method, radiographic image processing apparatus, radiographic image processing system, program, computer-readable storage medium, image diagnosis assisting method, and image diagnosis assisting system
US20030190064A1 (en) * 2002-04-03 2003-10-09 Hitoshi Inoue Radiographic image processing method, radiographic image processing apparatus, radiographic image processing system, program, computer-readable storage medium, image diagnosis assisting method, and image diagnosis assisting system
US20040109594A1 (en) * 2002-12-10 2004-06-10 Eastman Kodak Company Method for automatic construction of 2D statistical shape model for the lung regions
US20040109595A1 (en) * 2002-12-10 2004-06-10 Eastman Kodak Company Method for automated analysis of digital chest radiographs
US7221787B2 (en) * 2002-12-10 2007-05-22 Eastman Kodak Company Method for automated analysis of digital chest radiographs
US20070058850A1 (en) * 2002-12-10 2007-03-15 Hui Luo Method for automatic construction of 2d statistical shape model for the lung regions
US7221786B2 (en) * 2002-12-10 2007-05-22 Eastman Kodak Company Method for automatic construction of 2D statistical shape model for the lung regions
US20070086640A1 (en) * 2002-12-10 2007-04-19 Hui Luo Method for automated analysis of digital chest radiographs
US20050165294A1 (en) * 2003-03-18 2005-07-28 Weiss Kenneth L. Automated brain MRI and CT prescriptions in Talairach space
US7450983B2 (en) 2003-03-18 2008-11-11 University Of Cincinnati Automated brain MRI and CT prescriptions in Talairach space
US20050105788A1 (en) * 2003-11-19 2005-05-19 Matthew William Turek Methods and apparatus for processing image data to aid in detecting disease
US7346203B2 (en) 2003-11-19 2008-03-18 General Electric Company Methods and apparatus for processing image data to aid in detecting disease
US20050169536A1 (en) * 2004-01-30 2005-08-04 Vittorio Accomazzi System and method for applying active appearance models to image analysis
US8175894B2 (en) * 2004-03-09 2012-05-08 Kabushiki Kaisha Toshiba X-ray exposure report system, medical apparatus, and examination protocol distribution system
US20050209888A1 (en) * 2004-03-09 2005-09-22 Naoki Oowaki X-ray exposure report system, medical apparatus, and examination protocol distribution system
US20070014463A1 (en) * 2004-10-15 2007-01-18 Brigham And Women's Hospital Factor Analysis in Medical Imaging
US20060083415A1 (en) * 2004-10-15 2006-04-20 Georges El Fakhri Factor analysis in medical imaging
US8175355B2 (en) 2004-10-15 2012-05-08 The Brigham And Women's Hospital, Inc. Factor analysis in medical imaging
US7127095B2 (en) * 2004-10-15 2006-10-24 The Brigham And Women's Hospital, Inc. Factor analysis in medical imaging
US7519211B2 (en) 2004-10-15 2009-04-14 The Brigham And Women's Hospital, Inc. Factor analysis in medical imaging
US7734102B2 (en) 2005-05-11 2010-06-08 Optosecurity Inc. Method and system for screening cargo containers
US20080062262A1 (en) * 2005-05-11 2008-03-13 Luc Perron Apparatus, method and system for screening receptacles and persons
US7991242B2 (en) 2005-05-11 2011-08-02 Optosecurity Inc. Apparatus, method and system for screening receptacles and persons, having image distortion correction functionality
US20080205718A1 (en) * 2005-05-23 2008-08-28 Koninklijke Philips Electronics, N.V. Automated Organ Linking for Organ Model Placement
US7801348B2 (en) * 2005-07-18 2010-09-21 Analogic Corporation Method of and system for classifying objects using local distributions of multi-energy computed tomography images
US20070014472A1 (en) * 2005-07-18 2007-01-18 Zhengrong Ying Method of and system for classifying objects using local distributions of multi-energy computed tomography images
US20070081700A1 (en) * 2005-09-29 2007-04-12 General Electric Company Systems, methods and apparatus for creation of a database of images from categorical indices
US7929737B2 (en) 2005-09-29 2011-04-19 General Electric Company Method and system for automatically generating a disease severity index
US20070092864A1 (en) * 2005-09-30 2007-04-26 The University Of Iowa Research Foundation Treatment planning methods, devices and systems
US20070127802A1 (en) * 2005-12-05 2007-06-07 Siemens Corporate Research, Inc. Method and System for Automatic Lung Segmentation
US7756316B2 (en) 2005-12-05 2010-07-13 Siemens Medicals Solutions USA, Inc. Method and system for automatic lung segmentation
US20070242863A1 (en) * 2006-04-13 2007-10-18 Bernice Eland Hoppel Methods and Apparatus for Contouring at Least One Vessel
US8105242B2 (en) * 2006-04-19 2012-01-31 Dräger Medical GmbH Process and device for lung ventilation
US20070246047A1 (en) * 2006-04-19 2007-10-25 Draeger Medical Ag & Co. Kg Process and device for lung ventilation
US8503749B2 (en) 2006-05-03 2013-08-06 Ut-Battelle, Llc Method and system for the diagnosis of disease using retinal image content and an archive of diagnosed human patient data
US8243999B2 (en) * 2006-05-03 2012-08-14 Ut-Battelle, Llc Method and system for the diagnosis of disease using retinal image content and an archive of diagnosed human patient data
US20070258630A1 (en) * 2006-05-03 2007-11-08 Tobin Kenneth W Method and system for the diagnosis of disease using retinal image content and an archive of diagnosed human patient data
US7899232B2 (en) 2006-05-11 2011-03-01 Optosecurity Inc. Method and apparatus for providing threat image projection (TIP) in a luggage screening system, and luggage screening system implementing same
US8494210B2 (en) 2007-03-30 2013-07-23 Optosecurity Inc. User interface for use in security screening providing image enhancement capabilities and apparatus for implementing same
US20110255761A1 (en) * 2007-06-26 2011-10-20 University Of Rochester Method and system for detecting lung tumors and nodules
US7925653B2 (en) * 2008-02-27 2011-04-12 General Electric Company Method and system for accessing a group of objects in an electronic document
US20090216794A1 (en) * 2008-02-27 2009-08-27 General Electric Company Method and system for accessing a group of objects in an electronic document
US9042629B2 (en) * 2008-05-14 2015-05-26 Koninklijke Philips N.V. Image classification based on image segmentation
US20110222747A1 (en) * 2008-05-14 2011-09-15 Koninklijke Philips Electronics N.V. Image classification based on image segmentation
US20100228143A1 (en) * 2009-03-09 2010-09-09 Drager Medical Ag & Co. Kg Apparatus and method to determine functional lung characteristics
US8321007B2 (en) * 2009-03-09 2012-11-27 Dräger Medical GmbH Apparatus and method to determine functional lung characteristics
US8953858B2 (en) 2010-01-28 2015-02-10 Radlogics, Inc. Methods and systems for analyzing, prioritizing, visualizing, and reporting medical images
US8594402B2 (en) 2010-03-05 2013-11-26 Fujifilm Corporation Image diagnosis support apparatus, method and program
CN102243759A (en) * 2010-05-10 2011-11-16 东北大学 Three-dimensional lung vessel image segmentation method based on geometric deformation model
US8907949B2 (en) 2010-08-23 2014-12-09 Fujifilm Corporation Image display apparatus, method and program
US9892566B2 (en) 2011-02-07 2018-02-13 Fujifilm Corporation Image processing apparatus, method and program
EP2485194A2 (en) 2011-02-07 2012-08-08 Fujifilm Corporation Image processing apparatus, method and program
WO2012109658A2 (en) * 2011-02-11 2012-08-16 Emory University Systems, methods and computer readable storage mediums storing instructions for segmentation of medical images
WO2012109658A3 (en) * 2011-02-11 2012-10-18 Emory University Systems, methods and computer readable storage mediums storing instructions for segmentation of medical images
EP2490176A1 (en) 2011-02-15 2012-08-22 Fujifilm Corporation Surgery-assistance apparatus, method and program
US8805471B2 (en) 2011-02-15 2014-08-12 Fujifilm Corporation Surgery-assistance apparatus, method and program
US20140198892A1 (en) * 2011-03-22 2014-07-17 Keisuke Yamakawa X-ray ct device, and method
US9592021B2 (en) * 2011-03-22 2017-03-14 Hitachi, Ltd. X-ray CT device, and method
US8903147B2 (en) 2011-03-30 2014-12-02 Fujifilm Corporation Medical report generation apparatus, method and program
US9632206B2 (en) 2011-09-07 2017-04-25 Rapiscan Systems, Inc. X-ray inspection system that integrates manifest data with imaging/detection processing
US9117291B2 (en) 2012-04-19 2015-08-25 Fujifilm Corporation Image processing apparatus, image processing method, and non-transitory storage medium
US20140029817A1 (en) * 2012-07-26 2014-01-30 International Business Machines Corporation Tumor classification based on an analysis of a related ultrasonic attenuation map
US8781202B2 (en) * 2012-07-26 2014-07-15 International Business Machines Corporation Tumor classification based on an analysis of a related ultrasonic attenuation map
WO2014015424A1 (en) * 2012-07-26 2014-01-30 International Business Machines Corporation Tumor classification based on an analysis of a related ultrasonic attenuation map
US9668712B2 (en) * 2012-10-11 2017-06-06 Carestream Health, Inc. Method and system for quantitative imaging
US9427203B2 (en) * 2012-10-11 2016-08-30 Carestream Health, Inc. Method and system for quantitative imaging
US9204853B2 (en) * 2012-10-11 2015-12-08 Carestream Health, Inc. Method and system for quantitative imaging
CN104732086A (en) * 2015-03-23 2015-06-24 深圳市智影医疗科技有限公司 Computer-assisted disease detection system based on cloud computing
US20170039737A1 (en) * 2015-08-06 2017-02-09 Case Western Reserve University Decision support for disease characterization and treatment response with disease and peri-disease radiomics
US10004471B2 (en) * 2015-08-06 2018-06-26 Case Western Reserve University Decision support for disease characterization and treatment response with disease and peri-disease radiomics
US20170035381A1 (en) * 2015-08-06 2017-02-09 Case Western Reserve University Characterizing disease and treatment response with quantitative vessel tortuosity radiomics
CN105701799A (en) * 2015-12-31 2016-06-22 东软集团股份有限公司 Method and device for segmenting lung blood vessel in lung mask image

Also Published As

Publication number Publication date Type
EP1315125A2 (en) 2003-05-28 application
JP2003225231A (en) 2003-08-12 application
EP1315125B1 (en) 2008-06-04 grant
US7058210B2 (en) 2006-06-06 grant
EP1315125A3 (en) 2003-07-09 application
DE60226942D1 (en) 2008-07-17 grant
JP4310099B2 (en) 2009-08-05 grant

Similar Documents

Publication Publication Date Title
US5412563A (en) Gradient image segmentation method
US6553356B1 (en) Multi-view computer-assisted diagnosis
US7747055B1 (en) Virtual endoscopy with improved image segmentation and lesion detection
Li et al. Computerized radiographic mass detection. I. Lesion site selection by morphological enhancement and contextual segmentation
US7130457B2 (en) Systems and graphical user interface for analyzing body images
Yoshida et al. CAD techniques, challenges, andcontroversies in computed tomographic colonography
US6249594B1 (en) Autosegmentation/autocontouring system and method
Reeves et al. On measuring the change in size of pulmonary nodules
US7356367B2 (en) Computer aided treatment planning and visualization with image registration and fusion
Despotović et al. MRI segmentation of the human brain: challenges, methods, and applications
US7876939B2 (en) Medical imaging system for accurate measurement evaluation of changes in a target lesion
Lin et al. Computer-aided kidney segmentation on abdominal CT images
US6766043B2 (en) Pleural nodule detection from CT thoracic images
US20050207630A1 (en) Lung nodule detection and classification
Kostis et al. Three-dimensional segmentation and growth-rate estimation of small pulmonary nodules in helical CT images
US20100265251A1 (en) Virtual Endoscopy with Improved Image Segmentation and Lesion Detection
US7274810B2 (en) System and method for three-dimensional image rendering and analysis
Sluimer et al. Computer‐aided diagnosis in high resolution CT of the lungs
US7876938B2 (en) System and method for whole body landmark detection, segmentation and change quantification in digital images
US20050152588A1 (en) Method for virtual endoscopic visualization of the colon by shape-scale signatures, centerlining, and computerized detection of masses
US20060228015A1 (en) System and method for detection and display of diseases and abnormalities using confidence imaging
US7336809B2 (en) Segmentation in medical images
US20070116332A1 (en) Vessel segmentation using vesselness and edgeness
US20050110791A1 (en) Systems and methods for segmenting and displaying tubular vessels in volumetric imaging data
US20050281447A1 (en) System and method for detecting the aortic valve using a model-based segmentation technique

Legal Events

Date Code Title Description
AS Assignment

Owner name: GENERAL ELECTRIC COMPANY, NEW YORK

Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:MUNDY, JOSEPH LEAGRAND;MCCULLOCH, COLIN CRAIG;AVILA, RICARDO SCOTT;AND OTHERS;REEL/FRAME:012615/0473;SIGNING DATES FROM 20011119 TO 20020110

FPAY Fee payment

Year of fee payment: 4

REMI Maintenance fee reminder mailed
LAPS Lapse for failure to pay maintenance fees
FP Expired due to failure to pay maintenance fee

Effective date: 20140606